We propose a new concept called voice topography, referring to how upward voice behavior is distributed among team members. Building on “efficacy-safety” view of voice antecedents, we argue for two types of voice topography – status-based and LMX-based voice topography. Status- based voice topography refers to voice distribution within each team along team members’ status and LMX-based voice topography is defined as voice distribution within each team along team members’ LMX. Relying upon weak tie theory (Granovetter, 1973, 1983), we hypothesize that teams with high status-based or low LMX-based voice topography perform better. Data from 95 team leaders and 640 employees from 95 teams provided evidence that LMX-based voice topography was negatively related to team performance, but we did not find any significant effect of status-based voice topography.